https://github.com/mainanick/kerasplugins

kerasplugins: Boost Keras 💻
https://github.com/mainanick/kerasplugins

keras machine-learning python tensorflow

Last synced: 3 months ago
JSON representation

kerasplugins: Boost Keras 💻

• Host: GitHub
• URL: https://github.com/mainanick/kerasplugins
• Owner: mainanick
• License: mit
• Created: 2017-10-31T21:52:27.000Z (about 7 years ago)
• Default Branch: master
• Last Pushed: 2018-02-21T15:21:18.000Z (almost 7 years ago)
• Last Synced: 2024-10-01T16:08:18.185Z (3 months ago)
• Topics: keras, machine-learning, python, tensorflow
• Language: Python
• Homepage:
• Size: 27.3 KB
• Stars: 6
• Watchers: 3
• Forks: 1
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: LICENSE.txt

Awesome Lists containing this project

README

        
# kerasplugins 🤓

[![PyPI](https://img.shields.io/pypi/v/kerasplugins.svg)]()

[![PyPI](https://img.shields.io/pypi/l/kerasplugins.svg)]()

##### Installation

```

pip install kerasplugins

```

## Callbacks

### Notification Callback 📮 💬

Get notified as your Keras neural network models trains remotely

Send log messages to your telegram and slack channels. *more coming soon*

Notify levels available:

1) on_train_begin,

2) on_train_end,

3) on_batch_begin,

4) on_batch_end,

5) on_epoch_begin,

6) on_epoch_end

```python

from kerasplugins import callbacks

#Notify can either be a list, dict or set

notify = { 

  'on_batch_end',  # sends BATCH END: Loss 0.50 Accuracy: 0.75

  'on_epoch_end'   # sends EPOCH END: Loss 0.43 Accuracy: 0.81

}

# msg is the initial message

msg = "Predicting Bitcoin Price"

telegram = callbacks.TelegramNotify(, , msg=msg, notify=notify)

# channel is "#general" by default

slack = callbacks.SlackNotify(, , msg=msg, notify=notify)

webhook = callbacks.WebhookNotify('https://example.com/ml')

# Optionals:

#Headers default to "application/json"

headers = {'Content-Type': 'text/plain'}

# Added to every payload sent to the webhook

data = {"auth_token":"cccxxxvvvbbbnnnmmmaaasss"}

webhook = callbacks.WebhookNotify('https://example.com/ml', data=data, headers=headers)

model.fit(X_train, Y_train, validation_data=[X_test, Y_test], batch_size=256, epochs=10,

        callbacks=[telegram, slack, webhook])

```

## Coming Soon

1) Ability to stop training remotely